Better management of Failed Jobs

Would be great to be able to properly manage/handle failed jobs.

Currently when setting up the recipe, if there is a failed job, and it will never be a successful job because it shouldn't have been run (trigger or something wrong in the recipe), we would like to be able to delete it. 

Effectively 'managing' failed jobs, so that when looking at recipes, the number in red for failed jobs, are to be actioned alerts for us. Rather than something that can never be solved. 

Currently I have no way of marking that fail as something looked at and resolved. Instead its a red alert number forever glaring at me...

Hi there! Thanks for pointing this out, we appreciate your feedback! May I ask if red marking is not a good reminder that it is a failed jobs? Did you try to rerun the failed job when you have resolved it? The red color mark should disappear if the issue in the recipe is indeed resolved:) What do you think of adding more factors to the filer? 

Also being able to delete failed jobs would be great.  I'd like to not have any failed jobs and the only way to get rid of failed jobs is the copy the recipe and delete the old one.

I second Jeremy's suggestion. It would be great to clear out failed jobs and get a better idea of what's going on when a large job is run.

I couldn't agree more.  It's hard to keep it organized when you have errors showing that you don't need to rerun but also can't delete.

Something to add on this.  I've noticed that if you add error monitoring to the recipe that the job does not show up as failed.  I would change this as the job still failed but it completed the actions for the error monitoring.  Given that the only way to rerun a job is manual you still would need to take some action on the job.  When you use this feature you can't filter out failed jobs anymore.  Maybe giving the option on the Monitor for errors action would allow this to be used that way or not. 

Thanks for highlighting, Jeremy. 

Here are two recommendations that can help in such a situation.

1) If you would like to fail the job after the job is processed, you can process error handling logic in 'On error' (or not). Finally, Stop job with error after the monitor. 

The error catch within the error monitor can be configured to send an email or log a message including information about the failed job/error for your rectification. It's up to you to map the fields. You are right to say that the job will continue without failing, until it reaches the Stop job with error action.

Use case: You are processing a list of 100 items within the error monitor and an error is encountered on the 1st item. An error message is logged, and the job continues with the other 99 items, before failing at the end.

2) If you would like to fail the job immediately after an error is caught by the monitor, nest Stop job with error inside the 'On error' block.

Use case: You are processing a list of 100 items within the error monitor and an error is encountered on the 1st item. An error message is logged, and the job fails immediately.

===

Making use of Action with error monitors as well as Stop job with error actions allow you to pick up errors and rectify your recipes quickly. It also gives you better control over which jobs complete or fail by deciding how your recipe handle errors.

We hope this helps with your workflow. Have a go!

I agree with so many of the ideas in this thread. Error monitoring is a real pain with existing functionality, particularly when running lots of recipes and lots of jobs. If I can't stay on top of re-running every single job with an error (as many people point out is often resolved manually instead) then the red error count becomes useless.

I'd very much appreciate a centralized dashboard for Jobs rather than having to open each recipe individually to see if there are any "new" errors.

Alternatively, the count of recipe Jobs and Errors should be more dynamic or time-sensitive so that I can see what's happened over the past day/week/month rather than seeing the all time count, which is pretty useless.

any progress on this?

I can see the idea was raised 7 years ago. Managing failed jobs, especially the ones do not need to be re-run, becomes a pain for me now.

Is there a way that I can add a marker or a label for a failed job, such as ignore, no action. Those failed jobs are failed for testing purpose, they do not need to be re-executed. However, with many failed jobs, it's really hard to recognise which one needs to be resolved as an issue.

I typically like to see 0 Failed jobs. Having a any number indicates there is something that needs attention. Based on the recipe and what it was doing, there are times you cannot resolve the issue without it creating a new job leaving the old job failed.  Then after 30 days you cannot even see the details. It would be nice to have either the ability to clear the "failed" status so that the counter returns to 0 or after 30 days when the failed job details falls off, then so does the count.  That way you would only see the failed with details for 30 days.